superset-sh/superset

What it solves

Superset eliminates the context-switching overhead and interference that occurs when running multiple AI coding agents. It allows developers to run dozens or hundreds of agents in parallel, each isolated in its own environment, so they can compare results and merge the best solution without waiting for agents to run sequentially.

How it works

The platform manages CLI-based coding agents by deploying each one into an isolated git worktree. This ensures that every agent has its own branch, terminal, and environment. Superset provides a unified interface to monitor these agents, a built-in diff viewer for reviewing changes, and a terminal for interaction. It can be accessed via a desktop app, a CLI, a TypeScript SDK, or an MCP server, allowing for remote access and programmatic control.

Who it’s for

Developers who use CLI-based AI coding agents (such as Claude Code, Cursor Agent, or GitHub Copilot) and want to increase their productivity by running multiple tasks or experiments simultaneously.

Highlights

• Parallel Workspaces: Run 100+ agents simultaneously, each in an isolated git worktree.
• Agent Monitoring: Track agent status with working indicators and notifications.
• Built-in Tooling: Includes a terminal with rich prompt editing, a diff viewer for reviewing agent changes, and an in-app browser for previewing dev servers.
• Automations: Schedule agent sessions for recurring tasks like triaging issues or updating dependencies.
• Broad Compatibility: Works with any CLI-based agent, including Claude Code, OpenAI Codex, and Gemini CLI.
• Remote Access: Connect to and manage workspaces from different machines via the desktop app, CLI, or phone.